Disclaimer: I recently attended Storage Field Day 17. My flights, accommodation and other expenses were paid for by Tech Field Day. There is no requirement for me to blog about any of the content presented and I am not compensated in any way for my time at the event. Some materials presented were discussed under NDA and don’t form part of my blog posts, but could influence future discussions.
They maintain a strict focus on the SMB and Enterprise ROBO markets, and aren’t looking to be the next big thing in the enterprise any time soon.
So What’s All This About NVMe [over Fabrics]?
According to Max and the team, NVMe over Fabrics is “the next big thing in [network] storage”. Here’s a photo of Max saying just that.
Why Hate SAS?
It’s not that people hate SAS, it’s just that the SAS protocol was designed for disk, and NVMe was designed for Flash devices.
SAS (iSCSI / iSER)
NVMe [over Fabrics]
Complex driver built around archaic SCSI
Simple driver built around block device (R/W)
Single short queue per controller
One device = one controller, no bottlenecks
Single short queue per device
Many long queues per device
Serialised access, locks
Non-serialised access, no locks
Many-to-Many, true Point-to-Point
You Do You, Boo
StarWind have developed their own NVMe SPDK for Windows Server (as Intel doesn’t currently provide one). In early development they had some problems with high CPU overheads. CPU might be a “cheap resource”, but you still don’t want to use up 8 cores dishing out IO for a single device. They’ve managed to move a lot of the work to user space and cut down on core consumption. They’ve also built their own Linux (CentOS) based initiator for NVMe over Fabrics. They’ve developed a NVMe-oF initiator for Windows by combining a Linux initiator and stub driver in the hypervisor. “We found the elegant way to bring missing SPDK functionality to Windows Server: Run it in a VM with proper OS! First benefit – CPU is used more efficiently”. They’re looking to do something similar with ESXi in the very near future.
Thoughts And Further Reading
I like to think of StarWind as the little company from the Ukraine that can. They have a long, rich heritage in developing novel solutions to everyday storage problems in the data centre. They’re not necessarily trying to take over the world, but they’ve demonstrated before that they have an ability to deliver solutions that are unique (and sometimes pioneering) in the marketplace. They’ve spent a lot of time developing software storage solutions over the years, so it makes sense that they’d be interested to see what they could do with the latest storage protocols and devices. And if you’ve ever met Max and Anton (and the rest of their team), it makes even more sense that they wouldn’t necessarily wait around for Intel to release a Windows-based SPDK to see what type of performance they could get out of these fancy new Flash devices.
All of the big storage companies are coming out with various NVMe-based products, and a number are delivering NVMe over Fabrics solutions as well. There’s a whole lot of legacy storage that continues to dominate the enterprise and SMB storage markets, but I think it’s clear from presentations such as StarWind’s that the future is going to look a lot different in terms of the performance available to applications (both at the core and edge).
You can check out this primer on NVMe over Fabrics here, and the ratified 1.0a specification can be viewed here. Ray Lucchesi, as usual, does a much better job than I do of explaining things, and shares his thoughts here.
Disclaimer: I recently attended Storage Field Day 15. My flights, accommodation and other expenses were paid for by Tech Field Day. There is no requirement for me to blog about any of the content presented and I am not compensated in any way for my time at the event. Some materials presented were discussed under NDA and don’t form part of my blog posts, but could influence future discussions.
StarWind recently presented at Storage Field Day 15. You can see videos of their presentation here, and download my rough notes from here.
VTL? Say What Now?
Max and Anton from StarWind are my favourites. If I was a professional analyst I wouldn’t have favourites, but I do. Anyone who calls their presentation “From Dusk Till Dawn” is alright in my books. Here’s a shot of Max presenting.
In The Beginning
The concept of sending recovery data to tape is not a new one. After all, tape was often referred to as “backup’s best friend”. Capacity-wise it’s always been cheap compared to disk, and it’s been a (relatively) reliable medium to work with. This was certainly the case in the late 90s when I got my start in IT. Since then, though, disks have come a long way in terms of capacity (and reduced cost). StorageTek introduced Virtual Tape Libraries (VTLs) in the late 90s and a lot of people moved to using disk storage for their backups. Tape still played a big part in this workflow, with a lot of people being excited about disk to disk to tape (D2D2T) architectures in the early 2000s. IT was cool because it was a fast way to do backups (when it worked). StarWind call this the “dusk” of the VTL era.
Disks? Object? The Cloud? Heard Of Them?
According to StarWind though (and I have anecdotal evidence to support this), backup applications (early on) struggled to speak sensibly to disk. Since then, object storage has become much more popular. StarWind also suggested that it’s hard to do file or block to object effectively.
Tape (or a tape-like mechanism) for cold data is still a great option. No matter how you slice it, tape is still a lot cheaper than disk. At least in terms of raw $/GB. It also offers:
Can be stored offline; and
Streams at a reasonably high bandwidth.
Object storage is a key cloud technology. And object storage can deliver similar behaviour to tape, in that it is:
Capable of big IO; and
Doesn’t need random writes.
From StarWind’s perspective, the “dawn” of VTL is back. The combination of cheap disk, mature object storage technology and newer backup software means that VTL can be a compelling option for business that still needs a tape-like workflow. They offer a turnkey appliance, based on NL-SAS. It has 16 drives per appliance (in a 3.5” form factor), delivering roughly 120TB of capacity before deduplication. You can read more about it here.
Thoughts And Conclusion
StarWind never fail to deliver an interesting presentation at Tech Field Day events. I confess I didn’t expect to be having a conversation with someone about their VTL offering. But I must also confess that I do come across customers in my day job who still need to leverage VTL technologies to ensure their data protection workflow continues to work. Why don’t they re-tool their data protection architecture to get with the times? I wish it were that simple. Sometimes the easiest part of modernising your data protection environment is simply replacing the hardware.
StarWind are not aiming to compete in enterprise environments, focusing more on the SMB market. There are some nice integration points with their existing product offerings. And the ability to get the VTL data to a public cloud offering will keep CxOs playing the “cloud at all cost” game happy as well.
[Image courtesy of StarWind]
There are a lot of reasons to get your data protected in as many locations as possible. StarWind has a good story here with the on-premises part of the equation. According to StarWind, VTL will remain around “until backup applications (all of them) learn all cloud and on-premises object storage APIs … or until all object storage settles on a single, unified “standard” API”. This looks like it might still be some time away. A lot of environments are still using technology from last decade to perform business-critical functions inside their companies. There’s no shame in delivering products that can satisfy that market segment. It would be nice if everyone would refactor their applications for cloud, but it’s simply not the case right now. StarWind understand this, and understand that VTL is performs a useful function right now, particularly in environments where the advent of virtualisation might still be a recent event. I know people still using VTL in crusty mainframe environments and flashy, cloud-friendly, media and entertainment shops. Tape might be dead, but it feels like there are a lot of folks still using it, or its virtual counterpart.
Disclaimer: I recently attended Storage Field Day 12. My flights, accommodation and other expenses were paid for by Tech Field Day. There is no requirement for me to blog about any of the content presented and I am not compensated in any way for my time at the event. Some materials presented were discussed under NDA and don’t form part of my blog posts, but could influence future discussions.
Here are some notes from StarWind‘s presentation at Storage Field Day 12. You can view the video here and download my rough notes here.
StarWind Storage Appliance
StarWind’s Storage Appliance is looking to solve a number of problems with traditional storage solutions, including:
Unpredictable data and performance requirements growth;
Legacy storage incompatibility with existing compute / HCI; and
Untenable licensing costs when scaling out traditional HCI.
StarWind tell us that their solution offers the following benefits:
Easily scalable, fast, and fault tolerant storage;
Seamless integration into any existing CI/HCI; and
Ability to scale storage independently from compute.
This all sounds great, but what’s the “sweet spot” for deploying the appliances? StarWind position the solution where “high performance is needed for particular workload at a reasonable cost”. This seems to fall within the 20 – 40TB range. They also offer hybrid and all-flash models (you can check out the datasheet here).
What’s in the box?
So what do you get with these things? Nothing too surprising.
For StarWind, the real value is derived from the software platform driving the appliance.
Supported storage protocols include:
NVMoF (coming soon)
From a management perspective, you get access to the following tools:
VASA / vVols
While there’s support for Ethernet and InfiniBand, there is still (disappointingly) no FCoTR support.
Stairway to Cloud
StarWind also walked through their partner Aclouda‘s product – a hardware cloud storage gateway recognised by the server as an ordinary hard drive. In the picture above you can see:
SATA interface for host connectivity
Proprietary RTOS (no Linux!). Motorola PowerPC and ARM
Gigabit Ethernet: iSCSI and SMB3 for cloud uplink
Altera FGPA to accelerate (what software can’t do)
The idea is you can replace spinning disks with cloud storage transparently to any software-defined storage and hypervisor. You can read about the StarWind Storage Appliance and AcloudA Use Casehere (registration required).
Further Reading and Conclusion
I’m the first to admit that I was fairly ignorant of StarWind’s offering beyond some brief exposure to their free tools during a migration project I did a few years ago. Their approach to hardware seems solid, and they’ve always been a bit different in that they’ve traditionally used Windows as their core platform. I get the impression there’s a move away from this as scalability and throughput requirements increase at a rapid pace.
The HCI market is crowded to say the least, but this doesn’t mean companies like StarWind can’t deliver a reasonable product to customers. They say the sweet spot for this is 20 – 40TB, and there are plenty of smaller shops out there who’d be happy to look at this as an alternative to the bigger storage plays. To their credit, StarWind has focused on broad protocol support and useful management features. I think the genesis of the product in a software platform has certainly given them some experience in delivering features rather than relying on silicon to do the heavy lifting.
I’m looking forward to seeing how this plays out for StarWind, as I’m certainly keen to see them succeed (if for no other reason than Max and Anton are really nice guys). It remains to be seen whether the market is willing to take a bet on a relative newcomer to the HCI game, but StarWind appear to have the appetite to make the competition interesting, at least in the short term. And if you’ve gotten nothing else from this post, have a look at some of the white papers on the site, as they make for some great reading (registration required):
In what can only be considered excellent news, I’ll be heading to the US in early March for another Storage Field Day event. If you haven’t heard of the very excellent Tech Field Day events, you should check them out. I’m looking forward to time travel and spending time with some really smart people for a few days. It’s also worth checking back on the Storage Field Day 12 website during the event (March 8 – 10) as there’ll be video streaming and updated links to additional content. You can also see the list of delegates and event-related articles that have been published.
I think it’s a great line-up of presenting companies this time around. There are a few I’m very familiar with and some I’ve not seen in action before.
It’s not quite a total greybeard convention this time around, but I think that’s only because of Jon‘s relentless focus on personal grooming. I won’t do the delegate rundown, but having met a number of these people I can assure the videos will be worth watching.
Here’s the rough schedule (all times are ‘Merican Pacific and may change).
I’d like to publicly thank in advance the nice folks from Tech Field Day who’ve seen fit to have me back, as well as my employer for giving me time to attend these events. Also big thanks to the companies presenting. It’s going to be a lot of fun. Seriously.